Jobless Claims and History

Yesterday's post about the trendless trend in new jobless claims inspired one reader to complain that I wasn't paying sufficient attention to the history for this data series and was therefore drawing unnecessarily dark conclusions.

The charge is that I ignored the historical tendency for new filings for unemployment benefits to meander a bit in the wake of recessions past before turning down in earnest. After economic contractions end (as defined by the National Bureau of Economic Research), new weekly filings have been known to move sideways or even rise a bit for a time, providing false signals that the rebound is in trouble. The implication: the sluggish activity in filings this year should be viewed in the context of history. In other words, the trendless trend in 2010 isn't a troubling sign after all; rather, it's merely a repeat performance of the normal ebb and flow in post-recession periods.

That's a comforting thought, but how relevant is it for the here and now? Surely it would be foolish to dismiss history outright. And for the record, I'm no stranger to looking at the economic record and drawing inferences…up to a point. For example, in a March 2009 post, I looked at the history of initial jobless claims as an early signal for estimating when recessions will end. I've also noted at times that jobless claims don't always fall quickly after recessions (in an article from this past May, for instance). Do I point out all the caveats and possibilities in each and every post on initial jobless claims, or any other subject? Of course not--life (and attention spans) are short.

But I digress. Fo some context on the long-run history of claims, consider the chart below, which shows the seasonally adjusted four-week moving average of weekly new filings since the mid-1960s in comparison with recessions, as indicated by the gray bars. (Click on the chart to open a larger version.) Note that initial claims dropped once an economic downturn ended. No surprise there. But it's also true that the pace and magnitude of the drop varied considerably. This is economics, not physics. (Click to enlarge)

After the short recession in 1980, for instance, the four-week moving average peaked in early June of that year; six months later, it had fallen to just above 400,000. That was a stellar performance in terms of change: jobless claims that year fell rapidly and substantially. Of course, it didn't mean much that time--a new and deeper recession arrived the following year. By contrast, new claims fell sluggishly after the 1990-91 downturn. But the uninspiring decline didn't change the fact that the nineties would bring some of the strongest economic growth in U.S. history.

Every recession is different and so it's folly to expect history will repeat like clockwork when it comes to broad economic trends. Indeed, economic "rules" are in constant flux. One example can be found in the historical trend with job creation. In the wake of recessions over the past several decades, the net pace of job growth has been weakening after economic downturns. The fear is that this trend remains intact today. Given the deep losses in jobs, that trend threatens to bite deeper this time.

The past, in other words, isn't irrelevant for analyzing economic trends, but neither is it fate, depending on what we're looking at and when. Par for the course in the dismal science, which is also partly a dismal art.

The focus on these pages when it comes to initial jobless claims is trying to bring some perspective in real time to the numbers du jour. That's a difficult task, to say the least, and one that's prone to a fair amount of error, even under the best of circumstances. The future is still uncertain, no matter how many spreadsheets of data we review.

As to the optimistic view that the trendless trend so far this year with initial claims is typical and therefore nothing to worry about, we're skeptical. As the chart above reminds, claims rose to an unusually high level in the Great Recession and remains elevated. Even after more than a year of decline, claims are at heights that more or less represented peaks in the previous two recessions. History, it seems, cuts two ways when it comes to analyzing jobless claims. Pick your poison.

Granted, the last two downturns were mild compared to the recent troubles. But that only makes the prospect of a sluggish decline in this data series all the more troubling. It's reasonable to wonder if new jobless claims could move sideways for an unusually lengthy stretch of time, as they did after the 1990-91 contraction. Unless you're expecting a repeat of the late-1990s boom in the near-term future (we're not), the early nineties precedent for initial claims suggests a rocky road ahead.

Yes, there's a precedent for jobless claims staying relatively high, but that's hardly a reason for optimism given the current conditions, which are well short of inspirational. Dismissing the tendency for new claims to tread water this year based on looking to the past seems dangerously short-sighted at the moment. Perhaps our anxiety is misplaced; maybe salvation is coming after all. Surely there's good news in the economic numbers too, as we pointed out recently. But a mixed bag at this stage is problematic, as we've been discussing on various fronts all year.

Economic analysis is always enlightened until the next report changes the crowd's view. Meantime, the future is unclear and the case for staying cautious still has merit. History provides some incentive for keeping the faith, but it's just as easy to read the numbers from the past and draw darker visions.